Method and device for compressing an image and storage medium carrying an image compressed by said method

Abstract

The invention is related to a method for compressing images. The proposed method comprises associating perceptual importance parameters with pixels of the image, applying a transform on the image, partitioning the transformed image into code blocks and encoding coefficient bits of a given code block together from a most significant bit plane towards a least significant bit plane, wherein encoding of at least one coefficient is truncated at a truncation bit plane depending on the perceptual importance parameter associated with the pixel to which said code block coefficient corresponds. Truncating encoding of coefficients in dependency on the perceptual importance of the pixel associated with the coefficient allows for truncating coefficients corresponding to perceptually less important pixels at more significant bit planes. So, the overall perceptual quality of the compressed image is enhanced.

The invention claimed is: 
     
       1. A method for compressing an image, the method comprising
 associating perceptual importance parameters with pixels of the image, 
 applying a transform on the image, 
 partitioning the transformed image into code blocks and partitioning a given code block into sub-blocks, 
 using an average of the perceptual importance parameters of the pixels, to which the coefficients of the given code block correspond, for determining a preliminary truncation bit plane for said given code block, 
 encoding coefficient bits of said given code block together from a most significant bit plane towards a least significant bit plane, wherein 
 encoding of coefficient bits is truncated individually at sub-block dependent truncation bit planes, each of said sub-block dependent truncation bit planes being dependent on the perceptual importance parameters associated with pixels to which the coefficients of the respective sub-block correspond, wherein 
 the encoding of coefficients of at least one sub-block is truncated at an advanced truncation bit plane and encoding of coefficients comprised in at least another sub-block is truncated at a reduced truncation bit plane wherein an average truncation bit plane of the sub-blocks corresponds to said preliminary truncation bit plane. 
 
     
     
       2. The method according to  claim 1 , wherein
 coefficients which are comprised in the same sub-block are associated with the same perceptual importance parameter and 
 the at least one sub-block's coefficients are associated with a perceptual importance parameter indicating smaller perceptual importance than the code block parameter and 
 the at least another sub-block's coefficients are associated with a perceptual importance parameter indicating greater perceptual importance than the code block parameter. 
 
     
     
       3. The method according to  claim 2 , further comprising the steps of
 determining truncation bit plane variation parameters for the sub-blocks wherein 
 said truncation bit plane variation parameters depend on the perceptual importance parameter associated with the respective sub-block, the code block parameter and/or a maximum perceptual importance parameter within said given block. 
 
     
     
       4. The method according to  claim 1 , wherein determining the average truncation bit plane comprises
 further weighting a weighted mean distortion calculation over said given code block's coefficients by a feature weight associated with said given code block and 
 determining the average truncation bit plane based on the further weighted mean distortion calculation. 
 
     
     
       5. The method according to  claim 4 , wherein
 said feature weight associated with said given code block further depends on an average perceptual importance parameter and/or a maximum among the code block parameters. 
 
     
     
       6. The method according to  claim 1 , wherein
 the perceptual importance parameters are quantization parameter offsets. 
 
     
     
       7. The method according to  claim 6 , further comprising the step of,
 for a given pixel, determining the quantization parameter offset based on a quantization parameter offset step size and/or based on an indication of the number of features present in a macro block comprising said given pixel wherein 
 the features are related to skin, texture and luminance level. 
 
     
     
       8. The method according to  claim 7 , further comprising the step of determining the presence of the texture feature by
 parting the given code block into elementary blocks, 
 determining for each elementary block a mean square error, 
 determining the minimum among the determined mean square errors and 
 comparing the minimum with a threshold. 
 
     
     
       9. The method according to  claim 7 , further comprising the step of determining the presence of the luminance feature by
 determining an averaged luminance value of the macro block, 
 comparing the averaged luminance value with a lower threshold and/or 
 comparing the averaged luminance value with an upper threshold. 
 
     
     
       10. The method according to  claim 8 , further comprising the step of determining the presence of the luminance feature by
 determining an averaged luminance value of the macro block, 
 comparing the averaged luminance value with a lower threshold and/or 
 comparing the averaged luminance value with an upper threshold. 
 
     
     
       11. A device for compressing an image, the device comprising
 means for associating perceptual importance parameters with pixels of the image, 
 means for transforming the image, 
 means for partitioning the transformed image into code blocks and partitioning a given code block into sub-blocks, 
 means for using an average of the perceptual importance parameters of the pixels, to which the coefficients of the given code block correspond, for determining a preliminary truncation bit plane for said given code block, 
 encoding coefficient bits of said given code block together from a most significant bit plane towards a least significant bit plane, wherein 
 means for encoding of coefficient bits is truncated individually at sub-block dependent truncation bit planes, each of said sub-block dependent truncation bit planes being dependent on the perceptual importance parameters associated with pixels to which the coefficients of the respective sub-block correspond, wherein 
 said means for encoding being adapted such that the encoding of coefficients of at least one sub-block is truncated at an advanced truncation bit plane and encoding of coefficients comprised in at least another sub-block is truncated at a reduced truncation bit plane wherein an average truncation bit plane of the sub-blocks corresponds to said preliminary truncation bit plane.